Vulnerability
For this analysis, vulnerability refers to the susceptibility of individuals, households, or communities to adverse effects stemming from various interrelated factors, including economic, environmental, and social dimensions. This concept recognizes that vulnerability is not a singular issue but rather a complex interplay of multiple factors that can exacerbate risk and hinder resilience.
The methodology is based on widely used composite index construction techniques. A multidimensional vulnerability index is computed, combining various dimensions—drought vulnerability, health system vulnerability, mean nutrient adequacy, and per capita food expenditure—into a single, comprehensive measure. Each dimension is calculated as a sub-index from multiple indicators that represent relevant aspects of vulnerability in each domain. The indicators are based on a variety of nationally representative survey data including Malawi's Fifth Integrated Household Survey 2019/2020, Malawi's Household Panel Survey 2019, and Demographic and Health Survey (DHS) data, as well as satellite data.
- The Climate Change Vulnerability Index is based on (1) household survey data indicators on households’ exposure to climate risks, their sensitivity to these risks, and their adaptive capacity; and (2) satellite data on temperature and vegetation conditions. The satellite data is used to construct a drought index which uses a modified vegetation water supply index (MVWSI) defined as the ratio of the relative normalized difference vegetation index (RNDVI) and the square of relative land surface temperature (RLST). RNDVI and RLST are computed by dividing the observed values in the current year over the average values during the last 20 years.
- The Health System Vulnerability Index uses indicators on the share of women receiving assistance from a medical professional during their last childbirth and the share of women reporting that distance to a medical facility constitutes a major obstacle. Principal component analysis (PCA) is used to generate scores for each location for the strength of health systems.
- The Mean Nutrient Adequacy Ratio Index provides a global view of dietary quality by summarizing the adequacy of several nutrients simultaneously. It is calculated by averaging the individual Nutrient Adequacy Ratios (NAR) for a set of essential nutrients, where each NAR is determined by the ratio of an individual's daily nutrient intake to the Recommended Dietary Allowance (RDA) for that nutrient. To prevent over-consumption from artificially inflating the mean nutrient adequacy ratio, NAR values are capped at 1.
- The Per Capita Food Consumption Index reflects the total expenditure on food per person within a household or community. Lower per capita food expenditure is associated with higher vulnerability to food insecurity and malnutrition.
Following similar methodologies applied in previous vulnerability studies (Nkonde et al., 2014; Naudé et al., 2009), we normalize each indicator and sub-index using min-max normalization, transforming them into a common scale between 0 and 1, which facilitates comparability between the various indicators. The composite multidimensional vulnerability index is constructed by aggregating the sub-indexes from each dimension, using an unweighted mean which assumes equal importance to all four dimensions. Prior to aggregation, indicators are transformed so that higher values correspond to greater vulnerability (e.g., lower levels of access to healthcare and food security).
For each sub-indicator and for the composite vulnerability index, areas are classified as "much less", "less", "more", or "much more" vulnerable compared to the country's average. Thresholds between the four categories are designed such that approximately 25 percent of observations fall into each category, assuming a normal distribution. A higher value on the vulnerability index indicates greater vulnerability of a community to potential crises and complications relative to the national average.
This indicator is applied to assess vulnerability across Malawi’s districts. The results can help predict communities' capacity to absorb future shocks and inform response strategies. This information enables better targeting of preventive measures and facilitates more effective interventions to address the effects of crises when they impact communities.
Nutrient adequacy
The assessment of nutrient adequacy is based on two series of analysis. Firstly, we conduct an analysis of the dietary content of the food products consumed in order to identify areas with high nutrient intake gaps. Secondly, we apply a Quadratic Almost Ideal Demand System (QUAIDS) demand model to provide insights into the dynamics of household food consumption and derive nutrient income and price elasticities, which can be used to simulate the impacts of different policy options.
The analysis of nutrient intake is based on household consumption survey data from Malawi's Fifth Integrated Household Survey 2019/2020 and Malawi's Household Panel Survey 2019. We use the Malawian Food Composition Table for 2019 developed by the Government of Malawi (MAFOODS, 2019) to calculate energy and nutrient content of foods consumed, including calories, protein, iron, zinc, calcium, Vitamin A, folate (B9), thiamine (B1), riboflavin (B2), niacin (B3), Vitamin B6, Vitamin B12, and Vitamin C. To derive nutrient intake estimates for each household member, we apply the corresponding AME factors (FAO 2001) to account for household composition as we assume that food is distributed to members in proportion to each member's share of energy requirements. For each household, the household AME is calculated by summing the AMEs of individual household members. As consumption is recorded at the household level, the analysis does not account for potential unequal distribution of food within the household relative to nutrient needs. We also do not account for changes in energy and nutrient content due to food processing and cooking as we do not have information on how food is cooked by households. After calculation, observations outside the plausible consumption range of 500–5000 kcal consumption per person per day (Voortman et al. 2017) are dropped. Actual nutrient intakes are compared to recommended daily nutrient intakes defined in WHO/FAO (2005) to determine Nutrient Adequacy Ratios (NARs). As some households exceed recommended intakes and others have deficits, household NARs are truncated at 100 percent before calculating average consumption adequacy in order to avoid households with surplus intakes masking nutrient gaps in other households.
The elasticity analysis used a Quadratic Almost Ideal Demand System (QUAIDS) developed by Banks et al. (1997), which is the most appropriate form to derive income and price elasticities. It is a variant of the Lancaster (1971) model, which is able to capture the Engel curvature behavior. We first use the QUAIDS model to estimate price and income elasticities of major food groups. We then derive elasticities of nutrient demand with respect to income and food prices based on the estimated food demand elasticities and on food nutrient content. For further details on the nutrient adequacy analysis, see Magne Domgho et al. (2023).
References
Banks J., Blundell R., and Lewbel A. 1997. “Quadratic Engel Curves and Consumer Demand.”The Review of Economics and Statistics 79(4): 527–539.
FAO. 2001. Human Energy Requirements: Report of a Joint FAO/WHO/UNU Expert Consultation. Rome: Food and Agricultural Organization of the United Nations.
Lancaster K. 1971. Consumer Demand: A New Approach. Columbia University Press, New York.
MAFOODS. 2019. Malawian Food Composition Table, 1st Edition. Averalda van Graan, Joelaine Chetty, Malory Jumat, Sitilitha Masangwi, Agnes Mwangwela, Felix Pensulo Phiri, Lynne M. Ausman, Shibani Ghosh, Elizabeth Marino-Costello (Eds). Lilongwe, Malawi.
Magne Domgho L. J., Collins J., Ulimwengu, and O. Badiane. 2023. Identifying Nutrient Gaps and Priority Foods in Senegal. AKADEMIYA2063 Project Report Nutrient Smart Processing and Trade (NSPT) No. 002. Kigali: AKADEMIYA2063. https://doi.org/10.54067/nspt.002
Naudé, Wim, Amelia Santos-Paulino, and Mark McGillivray. 2009.“Measuring Vulnerability: An Overview and Introduction.”Oxford Development Studies 37 (3): 183–91.
Nkonde, M., M. B. Masuku, and A. M. Manyatsi. 2014. Factors Affecting Household Vulnerability to Climate Change in the Lowveld of Swaziland. FANRPAN Policy Brief. Pretoria: Food, Agriculture and Natural Resources Policy Analysis Network.
Voortman T., Kiefte-de Jong J. C., Ikram M. A., Stricker B. H., van Rooij F. J. A., Lahousse L., Tiemeier H., et al. 2017. “Adherence to the 2015 Dutch Dietary Guidelines and Risk of Non-Communicable Diseases and Mortality in the Rotterdam Study.” Eur J Epidemiol 32: 993–1005.
WHO/FAO. (2005). Vitamin and Mineral Requirements in Human Nutrition. 2nd edition. Geneva/Rome: World Health Organization, Food and Agricultural Organization of the United Nations.